By Anya Samek
We conducted a research study with a local grocery store to learn about customer food choices. We found that pre-ordering foods in advance increased the healthfulness of food purchased.
Many people have difficulty eating healthier, exercising more, saving more for the future, or quitting smoking, even when they say they want to make these changes. This is most apparent around February, when, after a burst of activity around early January, gyms become less busy again as people start abandoning their New Year’s resolutions.
Behavioral economists and psychologists theorize that self-control problems and temptation impulses are to blame for this phenomenon. That is, people want to make healthy changes, but have trouble sticking to them in the moment as temptations override their long-term plans. Behavioral economists call this deviation between advance choices (those that are made for the future) and immediate choices (those made in the moment) dynamic inconsistency.
In a study with colleagues at the University of California-San Diego (Sally Sadoff and Charles Sprenger), we measured dynamic inconsistency in a grocery store. Our study was conducted in partnership with Northgate Gonzalez Market in Los Angeles, CA (we previously ran a similar study with Louis’ Groceries in Chicago, IL). We offered over 200 shoppers in south-Central Los Angeles the option to pre-order a basket of foods from a large selection that contained fruits, vegetables, and sweet and salty snacks. When we delivered the baskets a week later, we surprised customers with the option to change their mind. We presented customers with all the options that were previously available, and customers could exchange up to 4 items.
Based on theories of temptation impulses, we hypothesized the following pattern: people would choose to exchange their cucumbers and carrots for chocolate bars and chips. This is exactly what we found. About 40% of our customers made at least one exchange, and the food people ended up with was significantly higher in calories, fat, and added sugar than the food people originally ordered. The remaining 60% of our customers did not make an exchange. This finding is an important contribution to science since we showed that in a real food context, about 40% of people are dynamically inconsistent – their advance and immediate choices differ in a predictable way. Our welfare estimates also indicates that welfare is reduced when people are inconsistent, so policies that mitigate inconsistency can improve outcomes.
We then asked: what can we do to mitigate the inconsistencies between advance and immediate choice? That is, how can we help people follow through with their advance choices? A common prescription in policy circles is to offer a “commitment device.” Commitment devices let people bind themselves to a particular decision. Pre-ordering groceries is a commitment device since we eliminate the impulsive purchases that happen at the grocery store.
We returned to our customers a week later and asked them to make another grocery pre-order for the following week. This time, before completing the order, we asked customers to tell us whether they’d like to stick to the choices they just made, or have the option to change their mind later. Choosing to stick to the choices they just made is evidence of wanting commitment, and we found that over half (56%) of our customers wanted the commitment option.
But it’s important for policy to know who wants to commit. If people who changed their mind in the first week commit, then pre-ordering food improves health outcomes. But if it is the people who didn’t change their mind in the first week who commit, then voluntary commitment won’t have a big impact. It turns out that the latter is true – people who previously changed their mind are less likely to choose commitment (43 percent) versus people who previously stuck to their initial choices (66 percent). This suggests that simply offering commitment devices isn’t going to improve food choice or welfare very much. But, our estimates suggest that making commitment mandatory (or, at least more attractive, through discounts and other incentives) would be more likely to improve food choice and welfare.
So the next time you go grocery shopping, think about doing it online. Pre-ordering through services like Instacart, Peapod, or directly through your grocery store is a commitment device that can help you avoid temptation impulses and follow-through with your New Year’s resolution at least a while longer.
This research is funded by the BECR Center at Duke University. We also want to thank Northgate Market for our excellent ongoing research partnership. Finally, we want to thank Andre Gray and the students at the Behavioral and Experimental Economics group for excellent research assistance.